Proceedings of the Society of Natural Sciences in Hamburg VERHANDLUNGEN. des Naturwissenschaftlichen Vereins in Hamburg NF GOECKE & EVERS

Transcription

1 Proceedings of the Society of Natural Sciences in Hamburg VERHANDLUNGEN des Naturwissenschaftlichen Vereins in Hamburg 46 NF GOECKE & EVERS

2 Verh. Naturwiss. Ver. Hamburg NF Page Simon Weigmann Hamburg Contribution to the taxonomy and distribution of eight ray species (Chondrichthyes, Batoidea) from coastal waters of Thailand Abstract A collection of 24 rays from eight different species, obtained in 1993 from the marine shelf around Thailand, was examined in this study. The rays were determined, morphometrically and meristically analyzed, photographically documented and compared with relevant literature and available further collection material. The tails of Dasyatis akajei were found to be distinctly longer than specified in literature. Compared to the references a larger maximal total length and morphometrical discrepancies for Rhinobatos formosensis as well as higher tooth row counts for Himantura gerrardi and Rhynchobatus australiae were detected. General taxonomical problems are discussed for the Rhynchobatidae. For Neotrygon kuhlii a much wider depth distribution than previously known was found. Furthermore, the known distribution area could be slightly extended for Himantura walga and strongly by about 2500 km for Rhinobatos formosensis. Author s Address Simon Weigmann, Biocenter Grindel and Zoological Museum, University of Hamburg, Martin-Luther-King- Platz 3, Hamburg, Germany. simon_weigmann@gmx.de

3 250 NF Simon Weigmann Introduction About 600 of the more than 1200 globally known Chondrichthyes species are rays (Last & Stevens 2009), with the highest diversity of cartilaginous fish living in the East Indian Ocean (Fowler et al. 2005, White et al. 2006, Ebert & Winton 2010). Due to the substantially increased targeted fishery, the populations of many elasmobranch species declined over the last decades. Further reasons for the drop in many elasmobranch species are bycatch, habitat loss and biotope degradation in combination with the late attainment of sexual maturity, slow growth and low reproductive output of most species (Camhi et al. 1998, Stevens et al. 2000, Myers & Worm 2003, Fowler et al. 2005). Thailand was the fifth important chondrichthyan fishing nation in the East Indian Ocean in 1997 with 5600 t officially landed (Vannuccini 1999). The examined specimens belong to the five Dasyatidae species Dasyatis akajei, D. zugei, Himantura gerrardi, H. walga and Neotrygon kuhlii, the Myliobatidae species Aetomylaeus nichofii, the Rhinobatidae species Rhinobatos formosensis and the Rhynchobatidae species Rhynchobatus australiae. The Dasyatidae is by far the most speciose batoid family in Thai and adjacent waters with 31 species (Vidthayanon 2002). Its members are characterized by one or more tail stings (missing only in the genus Urogymnus), the reduction to tail folds or absence of their caudal fins and whip-like tails (Last & Compagno 1999). The second speciose family in this region is the family Myliobatidae with only seven species (Vidthayanon 2002). In its members the anterior part of the head is extended beyond the disc and the eyes are located laterally on the sides of the head. Furthermore, the margin of the subrostral lobe is rounded and the floor of the mouth has fleshy papillae (Last & Stevens 2009). The Rhinobatidae with five and the Rhynchobatidae with four species in Thai and adjacent waters (Vidthayanon 2002) have a more or less shark-like body (Serena 2005), two well-developed and widely separated dorsal fins and no tail spine (Last & Stevens 2009). Furthermore, the members of both families have a caudal fin, which has a well-developed ventral lobe in the family Rhynchobatidae and a not well-defined ventral lobe in the Rhinobatidae (Last & Stevens 2009). Despite the numerous newly described elasmobranch species of the last years (e. g. Last et al. 2007, Last et al. 2008a, Last et al. 2008b, Last & Stevens 2009, Last et al. 2010), the biology and distribution of many species are still little known. In various cases the original descriptions of ray species are old and sketchy like those of Müller & Henle ( ) and comprehensive revisions or redescriptions have only rarely been made (Last & Stevens 2009). In several cases, the distinguishing features of very similar species from the same genus are not well-defined or based on outdated descriptions. In order to make a contribution to the filling of these knowledge gaps 24 ray speci-

4 Contribution to the taxonomy and distribution of eight ray species (Chondrichthyes, Batoidea) from coastal waters of Thailand 251 mens from coastal waters of Thailand were examined. This study provides extensive morphometrical analyses for eight ray species for most of which such detailed morphometrics have not been published before. Additionally, tooth row, vertebral and fin ray counts are given for all specimens as well as detailed morphological descriptions and comparisons with relevant literature. Maps showing the distribution areas to the full extent as well as radiographs are presented here for the first time for several species. Habitus photographs of adult female Himantura walga specimens with their uniquely thickened tails are also pictured for the first time. Furthermore, the depth distribution of Neotrygon kuhlii and the distribution areas of Himantura walga and Rhinobatos formosensis are extended. Material and Methods The examined batoids were collected by Matthias Stehmann during a Thailand expedition that took place from the 5 th to the 11 th December The specimens were collected from local fishermen in the four Thai harbors shown in the map in Fig. 1: Prachuap Khiri Khan (12 14 N 100 E), Chumphon (10 27 N E), Pak Phanang (8 20 N E) and Kantang (7 18 N E). The map was generated using ArcMap by ESRI (ESRI ) and based on the Global Relief Model ETOPO 1 of the Na- Figure 1 Map of Thailand showing the four harbors Prachuap Khiri Khan, Chumphon, Pak Phanang and Kantang in which batoids were acquired.

5 252 NF Simon Weigmann Figure 2 Schematic painting of a sting ray jaw after Bigelow & Schroeder (1953) showing how to count tooth rows in batoids. tional Geophysical Data Center (NOAA, Amante & Eakings 2009). The country borders were visualized by means of the shapefiles supplied by ESRI for the ArcExplorer-Java Edition for Education (AEJEE, ESRI ). Land below the sea level was colorized in the color of the lowest land elevation class using Adobe Photoshop CS 4(Adobe ). A total of 24 specimens was examined. All specimens were fixed in 4% formaldehyde solution soon after the catch and preserved in 70% ethanol afterwards. With regard to Himantura walga further material from the ZMH fish collection was examined for comparative purposes. Morphometrics and meristics were done following Bigelow & Schroeder (1953) with minor modifications unless otherwise stated. The tooth rows of the rays were counted in zigzag patterns as shown in Fig. 2 by means of a schematic painting of a sting ray jaw after Bigelow & Schroeder (1953), because the teeth of rays are arranged in a pavement pattern. This schematic jaw has 43 tooth rows in the upper and 41 rows in the lower jaw. For the Rhinobatidae and Rhynchobatidae additional measurements were taken after Norman (1926) and Last et al. (2004). Unless otherwise indicated the photographs were taken with a Nikon D90 and a Nikkor mm zoom lens and afterwards reworked using Adobe Photoshop CS4 (Adobe ). Radiographs of all examined specimens were taken with a 1979 launched MG 101 X-ray equipment for radiography by Philips. They were used for verifying the tooth row counts and for tak-

6 Contribution to the taxonomy and distribution of eight ray species (Chondrichthyes, Batoidea) from coastal waters of Thailand 253 ing internal meristics (vertebral and fin ray counts) of all specimens. The internal meristics were done following Krefft (1968) for the Myliobatiformes and after Compagno & Last (2008) for the Rhinobatiformes. Distribution maps are provided for all species with new distribution information or for which current distribution maps showing the distribution area to the full extend have not been available from other sources. The maps were generated using ArcMap (ESRI ) and based on the shapefiles supplied by ESRI for the ArcExplorer-Java Edition for Education (AE- JEE, ESRI ). The distribution areas and catch locations were drawn with Adobe Photoshop CS4 (Adobe ). Results The 24 examined ray specimens represent eight different species from two orders, four families and six genera. Their classifications, numbers of individuals and catch locations are shown in Table 1. Table 1 Classifications, numbers and locations of the examined specimens. class subclass order family genus species number harbor Chondrichthyes Elasmobranchii Myliobatiformes Dasyatidae Dasyatis akajei 2 Ch zugei 4 PK Himantura gerrardi 2 Kt walga 6 PK Neotrygon kuhlii 5 Kt Myliobatidae Aetomylaeus nichofii 1 Kt Rhinobatiformes Rhinobatidae Rhinobatos formosensis 2?? Rhynchobatidae Rhynchobatus australiae 2 PP

7 254 NF Simon Weigmann The abbreviations of the harbors stand for Ch = Chumphon, Kt = Kantang, PK = Prachuap Khiri Khan, PP = Pak Phanang,?? = exact Thai harbor unknown. Typical characteristics which proved to be important for the determination are provided for all specimens as well as comparisons with relevant literature and in the case of more complex determination procedures differences to similar species. Furthermore, comments about aberrations in the examined specimens from the descriptions in literature and possible mistakes or problems in the references including taxonomically problematic cases are given. These themes are not part of the conclusion chapter, but have been included directly in the results chapter to allow direct comparisons with the species descriptions. With the exception of the two specimens of Rhynchobatus australiae, which are pictured with three habitus photographs, two habitus photographs are shown for each of the examined specimens. Distribution maps are given for Dasyatis akajei, D. zugei, Himantura gerrardi, H. walga and Rhinobatos formosensis because no current maps illustrating the complete distribution areas of these species have been available. For distribution maps of the other examined species see for example Last & Stevens (2009). Tables with the measurements of all 24 analyzed specimens can be found in the appendix. A collection of sharks from the same expedition will be described by the author in a subsequent paper. Dasyatis akajei (Müller & Henle, 1841) The two specimens of Dasyatis akajei (ZMH and ZMH 25686) were caught by local fishermen in about 30 m depth in the Gulf of Thailand near Chumphon on the 6 th December As described for this species by Last & Compagno (1999) they have a short snout, no dark transverse band through the eyes, their ventral skin fold terminates anterior to the tip of the undamaged tail, their dorsal surface lacks color patterns and their tail is not banded posterior to the sting (Fig. 3, A D). Furthermore, the undersurface of the disc and their pelvic fins are whitish with a broad yellowish brown margin (Fig. 3, B, D). The two examined specimens differ from the similar species Dasyatis bennettii (Müller & Henle, 1841) in having a dorsal skin fold on their tail, which is not present in D. bennettii. Additionally, both specimens have many minute tubercles around the dorsal median row of thorns as well as five (ZMH 25686, Fig. 3, A) to six (ZMH 25685, Fig. 3, C) large thorns on the tail anterior to the sting. Dasyatis bennettii in contrast has only few and small thorns anterior to the sting and no tubercles around the median row (Last & Compagno 1999). In the similar species Dasyatis fluviorum Ogilby, 1908 and D. parvonigra Last & White, 2008a, which like D. akajei have a dorsal skin fold on their tail, the dorsal thorns are restricted to a median row of small thorns with no thorns on the tail (Last & Stevens 2009).

8 Contribution to the taxonomy and distribution of eight ray species (Chondrichthyes, Batoidea) from coastal waters of Thailand 255 Figure 3 Dasyatis akajei: A B: ZMH 25685: A: Dorsal view, B: Ventral view. Scale bar A B 5 cm. C D: ZMH 25686: C: Dorsal view, D: Ventral view. Scale bar C D 5 cm.

9 256 NF Simon Weigmann Both examined specimens of Dasyatis akajei differ from the references in having a tail that is more than two times longer than their disc width, while it is less than two times longer following Last & Compagno (1999). However, they also mention that only little is known about Dasyatis akajei in general. Following Müller & Henle (1841) this character seems to be very variable as in the seven specimens examined by them the tail was between one quarter to two times longer than the disc width. In the seven specimens analyzed by Müller & Henle (1841) it is most likely that at least in some of them the tails were partially broken and healed afterwards during their lifetimes. As this happens often in the whip-like tails of sting rays and is only hardly detectable in many cases, length measurements should always be utilized cautiously. Last & Stevens (2009), too, mentioned that length measurements are not always useful when describing the size of sting rays because of the often damaged whip-like tails. Further material would be necessary to do more comparisons of the tail length. The specimen ZMH has 42 tooth rows in the upper and 53 in the lower jaw, specimen ZMH has 41 in the upper and 51 rows in the lower jaw. The internal meristics are as follows: ZMH 25685: Vertebral centra anterior to the first sting: 97, total vertebral centra count: 118, monospondylous or pretransitional centra (from first complete centrum in synarcual to mono-diplospondyly transition): Figure 4 Dasyatis akajei: Radiograph of specimen ZMH

10 Contribution to the taxonomy and distribution of eight ray species (Chondrichthyes, Batoidea) from coastal waters of Thailand 257 Figure 5 Dasyatis akajei: Distribution area after Last & Compagno (1999) and GBIF (2010), with the catch location of specimens ZMH and ZMH marked as a blue spot. 37, diplospondylous centra: 81; pectoral fin rays (left/right): 111/111, ventral fin rays (left/right): 20/22 (excluding claspers). ZMH (Fig. 4): Vertebral centra anterior to the first sting: 104, total vertebral centra count: 119, monospondylous centra: 38, diplospondylous centra: 81; pectoral fin rays (left/right): 113/112, ventral fin rays (left/ right): 23/23 (excluding claspers). According to Last & Compagno (1999) Dasyatis akajei reaches a maximal disc width of at least 66 cm. Both in this study examined specimens already have at disc widths of 35.5 cm (ZMH 25685) and 32.3 cm (ZMH 25686) fully developed claspers (Fig. 3, A D) and thus can be considered to be adult. Therefore, this species probably does not grow much larger than 66cm disc width. Following Last & Compagno (1999) Dasyatis akajei is distributed from East Thailand to South Japan. Two habitus photographs of each of the specimens ZMH and ZMH are shown in Fig. 3 and a radiograph of specimen ZMH in Fig. 4. A distribution map for Dasyatis akajei is pictured in Fig. 5, in which the distribution area after Last & Compagno (1999) as well as the species dataset of the Global Biodiversity Information Facility (GBIF 2010) is marked in red and white stripes and the catch location of specimens ZMH and ZMH as a blue spot. Their measurements can be found in Tables A1 and A2.

11 258 NF Simon Weigmann Dasyatis zugei (Müller & Henle, 1841) The two male and two female specimens of Dasyatis zugei (ZMH 25689) were caught by local fishermen in the Gulf of Thailand near Prachuap Khiri Khan in about 50 m depth on the 5 th December Following Last & Compagno (1999) they have contrary to all other species of the family Dasyatidae in this region an extremely elongated snout with the anterior margin of the disc being broadly concave. As described by Last & Compagno (1999) for Dasyatis zugei they have no dark transverse band through the eyes, their ventral skin fold terminates anterior to the tip of the undamaged tail, their dorsal surface lacks color patterns and their tail is not banded posterior to the sting (Figs. 6 7). The small female with 210 mm disc width has 48 tooth rows in the upper and 45 in the lower jaw, the larger female with 233 mm disc width has 40 and 46 rows, the small male with 164 mm disc width has 41 and 43 and the larger male with 178 mm disc width has 45 and 41 tooth rows. Krefft (1961) mentions three juvenile specimens of Dasyatis zugei with 44 to 48 tooth rows in the upper and 45 to 52 rows in the lower jaw. The internal meristics are as follows: Female, 210 mm disc width (Fig. 8): Vertebral centra anterior to the first sting: 85, total vertebral centra count: 95, monospondylous centra: 37, diplospondylous centra: 58; pectoral fin rays (left/right): 111/111, ventral fin rays (left/right): 20/20. Female, 233 mm disc width: Vertebral centra anterior to the first sting: 89, total vertebral centra count: 95, monospondylous centra: 34, diplospondylous centra: 61; pectoral fin rays (left/right): 113/113, ventral fin rays (left/right): 21/19. Male, 164 mm disc width: Vertebral centra anterior to the first sting: 85, total vertebral centra count: 94, monospondylous centra: 36, diplospondylous centra: 58; pectoral fin rays (left/right): 108/109, ventral fin rays (left/right): 16/16 (excluding claspers). Male, 178 mm disc width: Vertebral centra anterior to the first sting: 80, total vertebral centra count: 86, monospondylous centra: 32, diplospondylous centra: 54; pectoral fin rays (left/right): 94/101, ventral fin rays (left/right): 15/14 (excluding claspers). According to Last & Compagno (1999) Dasyatis zugei reaches a maximal disc width of at least 29 cm and the commercial width averages about 18 cm. For this reason the two females can be considered as large specimens. The already fully developed claspers of the only 16.4 cm and 17.8 cm wide males (Fig. 7, A D) suggest the conclusion that this species does not grow much larger than 29 cm. Although only little is known about the biology of Dasyatis zugei, which is distributed from India over Southeast Asia to South Japan, this species is taken in commercial quantities in the Gulf of Thailand (Last & Compagno 1999). Two habitus photographs of each of the four specimens (ZMH 25689) are shown in Figs. 6 and 7 and a radiograph of the 210mm wide female in Fig. 8. A distribution

12 Contribution to the taxonomy and distribution of eight ray species (Chondrichthyes, Batoidea) from coastal waters of Thailand 259 Figure 6 Dasyatis zugei, ZMH 25689: A B: Female, 210 mm disc width: A: Dorsal view, B: Ventral view. Scale bar A B 5 cm. C D: Female, 233 mm disc width: C: Dorsal view, D: Ventral view. Scale bar C D 5 cm.

13 260 NF Simon Weigmann Figure 7 Dasyatis zugei, ZMH 25689: A B: Male, 164 mm disc width: A: Dorsal view, B: Ventral view. Scale bar A B 5 cm. C D: Male, 178 mm disc width: C: Dorsal view, D: Ventral view. Scale bar C D 5 cm.

14 Contribution to the taxonomy and distribution of eight ray species (Chondrichthyes, Batoidea) from coastal waters of Thailand 261 Figure 8 Dasyatis zugei, ZMH 25689: Radiograph of the 210 mm wide female. Figure 9 Dasyatis zugei: Distribution area after Last & Compagno (1999) and GBIF (2010), with the catch location of the four examined specimens (ZMH 25689) marked as a blue spot.

15 262 NF Simon Weigmann map for Dasyatis zugei is pictured in Fig. 9, in which the distribution area after Last & Compagno (1999) as well as the species dataset of the Global Biodiversity Information Facility (GBIF 2010) is marked in red and white stripes and the catch location of the four examined specimens as a blue spot. Their measurements can be found in Tables A3 A6. Himantura gerrardi (Gray, 1851) The two specimens of Himantura gerrardi (ZMH and ZMH 25692) were caught by local fishermen in about 40 m depth in the East Andaman Sea near Kantang on the 8 th December The most obvious morphological characters for the determination following Last & Compagno (1999) are a prominent greenish, pearl-like thorn dorsomedially, white spots on the upper surface and a very long and slender tail with alternating black and white bands on the upper half of the tail beyond the sting (Fig. 10, A D). As described by Last & Compagno (1999) both specimens have no complex color patterns on the upper surface besides the white spots. In specimen ZMH the white spots are restricted to the posterior part of the disc, the tail base and the pelvic fins (Fig. 10, A), whereas they are spread almost over the whole dorsal surface, the tail base and the pelvic fins in specimen ZMH (Fig. 10, C). Specimen ZMH has 36 tooth rows in the upper and 44 in the lower jaw, exemplar ZMH has 40 and 50 tooth rows. Krefft (1961) mentions three juvenile specimens of Himantura gerrardi with 33 to 37 tooth rows in the upper and 36 to 38 rows in the lower jaw, whereas Fowler (1941) lists only 13 tooth rows for the upper and 23 rows for the lower jaw of this species. Krefft (1961) recounted the teeth of the specimen analyzed by Fowler (1941) and got out 30 tooth rows in the upper and 28 rows in the lower jaw. He also redetermined this specimen as Himantura fai Jordan & Seale, 1906 instead of H. gerrardi. The internal meristics are as follows: ZMH (Fig. 11): Vertebral centra anterior to the first sting: 118, total vertebral centra count: 118, monospondylous centra: 50, diplospondylous centra: 68; pectoral fin rays (left/right): 140/140, ventral fin rays (left/right): 23/23 (excluding claspers). ZMH 25692: Vertebral centra anterior to the first sting: 112, total vertebral centra count: 112, monospondylous centra: 51, diplospondylous centra: 61; pectoral fin rays (left/right): 139/139, ventral fin rays (left/ right): 29/29. The free vertebral centra terminate far anterior to the first sting in both specimens of Himantura gerrardi (Fig. 11), whereas they terminate posterior to the sting in all other in this study examined species of stingray. Compared to the maximal disc width of 90 cm (Last & Compagno 1999), both specimens are clearly juveniles with 22 cm (ZMH 25691) and 31.5 cm (ZMH 25692), which is supported by the not fully developed claspers of the male ZMH (Fig. 10, B). According to Last & Compagno (1999) Himantura gerrardi lives on the continental

16 Contribution to the taxonomy and distribution of eight ray species (Chondrichthyes, Batoidea) from coastal waters of Thailand 263 Figure 10 Himantura gerrardi: A B: ZMH 25691: A: Dorsal view, B: Ventral view. Scale bar A B 5 cm. C D: ZMH 25692: C: Dorsal view, D: Ventral view. Scale bar C D 5 cm.

17 264 NF Simon Weigmann Figure 11 Himantura gerrardi: Radiograph of specimen ZMH Figure 12 Himantura gerrardi: Distribution area after Last & Compagno (1999) and GBIF (2010), with the catch location of specimens ZMH and ZMH marked as a blue spot.

18 Contribution to the taxonomy and distribution of eight ray species (Chondrichthyes, Batoidea) from coastal waters of Thailand 265 shelf with the exact depth distribution still being unknown. It is the commercially most important sting ray species in its distribution area, which ranges from India over Southeast Asia to Papua New Guinea in the East and Taiwan in the North. Records from the East and Southeast African coast need to be verified (Last & Compagno 1999). Two habitus photographs of each of the specimens ZMH and ZMH are shown in Fig. 10 and a radiograph of specimen ZMH in Fig. 11. A distribution map for Himantura gerrardi is pictured in Fig. 12, in which the distribution area after Last & Compagno (1999) as well as the species dataset of the Global Biodiversity Information Facility (GBIF 2010) is marked in red and white stripes and the catch location of specimens ZMH and ZMH as a blue spot. Their measurements can be found in Tables A7 and A8. Himantura walga (Müller & Henle, 1841) The three male and three female specimens of Himantura walga (ZMH 25690) were caught by local fishermen in the Gulf of Thailand near Prachuap Khiri Khan in about 50 m depth on the 5 th December As described for this species by Last & Compagno (1999) all six specimens have a broad dense band of low, flat and heart-shaped denticles from just anterior to the orbit over the middle of the disc to the tail spine (Figs , A, C). Furthermore, they are small-sized, have an acute snout with a prominent tip, an almost oval disc, no cutaneous folds as well as a plain brownish upper and a uniformly whitish ventral surface (Figs ). The short tail does not bear lateral keels. In the male specimens the tail is filamentous beyond the sting (Figs. 13, A D and 14, A B), but in the females it is shaped like a green bean, being initially bulbous and then terminating in a fine filament (Figs. 14, C D and 15, A D). Four Himantura walga specimens (ZMH ) were used for comparison (Fig. 16, A D). These specimens were found aboard a local fishing boat in Pulau Pangkor, West Malaysia on 19 th November They had been caught by the fishermen in m depth with a trawl net. Although the highly unusual transformation of the tail in mature females of Himantura walga was already described by Last & Compagno (1999), habitus photographs of adult females are pictured here for the first time (Figs. 14, C D and 15, A D). The function of the swelling is still unknown. The filamentous tails of the 79 mm and 93 mm wide juvenile females from the comparison material (ZMH , Fig. 16, C D) prove that the transformation really takes place not until the females get mature. Further evidence is supplied by the with 167 mm disc width smallest of the three adult females of ZMH 25690, in which the tail is only slightly thickened, but already shows the typically bent terminal filament (Fig. 14, C D). As both larger females with 181 mm and 185 mm disc width have fully thickened tails (Fig. 15, A D), females seem to

19 266 NF Simon Weigmann Figure 13 Himantura walga, ZMH 25690: A B: Male, 172 mm disc width: A: Dorsal view, B: Ventral view. Scale bar A B 5 cm. C D: Male, 174 mm disc width: C: Dorsal view, D: Ventral view. Scale bar C D 5 cm.

20 Contribution to the taxonomy and distribution of eight ray species (Chondrichthyes, Batoidea) from coastal waters of Thailand 267 Figure 14 Himantura walga, ZMH 25690: A B: Male, 187 mm disc width: A: Dorsal view, B: Ventral view. Scale bar A B 5 cm. C D: Female, 167 mm disc width: C: Dorsal view, D: Ventral view. Scale bar C D 5 cm.

21 268 NF Simon Weigmann Figure 15 Himantura walga, ZMH 25690: A B: Female, 181 mm disc width: A: Dorsal view, B: Ventral view. Scale bar A B 5 cm. C D: Female, 185 mm disc width: C: Dorsal view, D: Ventral view. Scale bar C D 5 cm.

22 Contribution to the taxonomy and distribution of eight ray species (Chondrichthyes, Batoidea) from coastal waters of Thailand 269 Figure 16 Himantura walga, ZMH , ventral views: A: Male, 112 mm disc width, B: Male, 110 mm disc width, C: Female, 93 mm disc width, D: Female, 79 mm disc width. Scale bar A D 5 cm. Camera: Sony Cyber-shot DSC W55. get mature at disc widths of about 170 mm. This assumption corresponds with the observations by White et al. (2006) and White & Dharmadi (2007). According to these authors the males of Himantura walga have a similar maturity size. All three males of ZMH with disc widths of 172 mm, 174 mm and 187 mm are mature with fully developed claspers (Figs. 13, B, D and 14, B), whereas the two males of the comparison material (ZMH ) are still juvenile with not fully developed claspers at widths of 110 mm and 112 mm (Fig. 16, A B). Tooth row counts for the six specimens from Thailand are as follows: Male, 172 mm wide: 39 tooth rows in the upper and 43 in the lower jaw; male, 174 mm: 45 and 48 tooth rows; male, 187 mm: 40 and 46 rows; female, 167 mm: 38 and 42 rows; female, 181 mm: 40 and 44 rows; female, 185 mm: 41 and 45 rows. The internal meristics are as follows: Male, 172 mm disc width (Fig. 17): Vertebral centra anterior to the first sting: 78, total vertebral centra count: ca. 90, monospondylous centra: 35, diplospondylous centra: ca. 55; pectoral fin rays (left/right): 101/100, ventral fin rays (left/right): 18/17 (excluding claspers). Male, 174 mm disc width: Vertebral centra anterior to the first sting: 78, total vertebral centra count: ca. 85, monospondylous centra: 36, diplospondylous centra: ca. 49; pectoral fin rays (left/right): 105/106, ventral fin rays (left/right): 17/17 (excluding claspers). Male, 187 mm disc width: Vertebral centra anterior to the first sting: 80, total vertebral centra count: ca. 90, monospondylous centra: 38, diplospondylous centra: ca. 52; pectoral fin rays (left/ right): 102/104, ventral fin rays (left/right): 18/17 (excluding claspers). Female, 167 mm disc width: Vertebral centra anterior to the first sting: 77, total vertebral centra count: ca. 87, monospondylous centra: 35, diplospondylous centra: ca. 52; pectoral fin rays (left/right): 103/102, ventral fin rays (left/right): 24/24. Female, 181 mm disc width (Fig. 18): Vertebral centra anterior to the first sting: 81, total vertebral centra count: 87, monospondylous centra: 38, diplospondylous centra: 49; pectoral fin rays (left/right): 102/102, ventral fin rays (left/right): 23/23. Female, 185 mm disc width:

23 270 NF Simon Weigmann Figure 17 Himantura walga, ZMH 25690: Radiograph of the 172 mm wide male. Vertebral centra anterior to the first sting: 83, total vertebral centra count: 90, monospondylous centra: 38, diplospondylous centra: 52; pectoral fin rays (left/right): 109/ 108, ventral fin rays (left/right): 23/24. Although Last & Compagno (1999) mention a maximal disc width of 18 cm for Himantura walga, the two larger females and the largest male of ZMH are slightly wider with 181 mm, 185 mm and 187 mm disc width. However, the maximal disc width of Himantura walga is about 24 cm following White et al. (2006) and White & Dharmadi (2007). According to these authors the birth size is about 8 to 11 cm disc width in Himantura walga. The 79 mm and the 93 mm wide females of the comparison material (ZMH ) represent the smallest known neonates of Himantura walga. These specimens indicate a smaller birth size for Himantura walga than previously assumed. The formerly smallest known neonate of this species was a 107 mm wide specimen reported by White & Dharmadi (2007). According to Last & Compagno (1999) little is known about the biology of this mainly, inshore living species, which is commercially fished in the Gulf of Thailand. The so far known distribution of Himantura walga ranges from the East coast of Thailand to Southeast Indonesia. Due to confusion with Himantura imbricata (Bloch & Schneider, 1801), the westward distribution of H. walga is still unclear (Last & Com-

24 Contribution to the taxonomy and distribution of eight ray species (Chondrichthyes, Batoidea) from coastal waters of Thailand 271 Figure 18 Himantura walga, ZMH 25690: Radiograph of the 181 mm wide female. Figure 19 Himantura walga: Distribution area after Last & Compagno (1999) and GBIF (2010), with the catch location of the six Thai specimens (ZMH 25690) marked as a blue spot and that of the four Malaysian specimens (ZMH ) as a black spot.

25 272 NF Simon Weigmann pagno 1999). The comparison material (ZMH ) was caught outside the so far known distribution area and represents the first confirmed record of this species in the Andaman Sea. Two habitus photographs of each of the six specimens (ZMH 25690) are shown in Figs and radiographs of the 172 mm wide male and the 181 mm wide female in Figs. 17 and 18. A distribution map for Himantura walga is pictured in Fig. 19, in which the distribution area after Last & Compagno (1999) as well as the species dataset of the Global Biodiversity Information Facility (GBIF 2010) is marked in red and white stripes, the catch location of the six Thai specimens (ZMH 25690) as a blue spot and the location of the four Malaysian individuals (ZMH ) as a black spot. The measurements of the Thai specimens can be found in Tables A9 A14. Neotrygon kuhlii (Müller & Henle, 1841) The five specimens of Neotrygon kuhlii (ZMH ) were caught by local fishermen in the East Andaman Sea near Kantang on the 8 th December The two embryos (ZMH 25700) were taken out of the 35.3 cm wide female ZMH As described by Last & Stevens (2009) the five specimens have like all species of the genus Neotrygon a dark band through the eyes, a short tail with alternating black and white bands on the upper half of the tail beyond the sting, a relatively small adult size and a well-developed ventral skin fold on their tail (Figs ). As specified by Last & Stevens (2009) and contrary to all other species of the genus the examined specimens except for the two embryos have prominent blue spots on their dorsal surface, which became discolored during the storage in alcohol (Figs. 20, A, C and 21, A). Despite this discolorization the determination by means of the spots is still unambiguous because no other species of the genus has similar spots, but, instead, rather complex color patterns like Neotrygon leylandi (Last, 1987) or small speckles like Neotrygon picta Last & White, 2008b. Further characters of the five examined specimens, as described for Neotrygon kuhlii by Last & Stevens (2009), are a rhomboidal disc, a thickened trunk, small denticles, which are confined to a single row along the disc midline and a welldefined dorsal and ventral skin fold on the tail (Figs ). Tooth row counts for the five examined specimens are as follows: ZMH 25697: 32 rows in the upper and 30 rows in the lower jaw, ZMH 25698: 36 and 39 rows, ZMH 25699: 29 and 32 rows, ZMH 25700: Embryo 1, 94 mm wide: 26 and about 30 rows; embryo 2, 95 mm wide: 31 and about 30 rows. The internal meristics are as follows: ZMH 25697: Vertebral centra anterior to the first sting: 105, total vertebral centra count: ca. 145, monospondylous centra: 44, diplospondylous centra: ca. 101; pectoral fin rays (left/right): 109/109, ventral fin rays (left/right): 20/20 (excluding claspers). ZMH (Fig. 22): Vertebral centra anterior to the first sting: 103, total vertebral centra count: ca. 142, monospondylous centra:

26 Contribution to the taxonomy and distribution of eight ray species (Chondrichthyes, Batoidea) from coastal waters of Thailand 273 Figure 20 Neotrygon kuhlii: A B: ZMH 25697: A: Dorsal view, B: Ventral view. Scale bar A B 5 cm. C D: ZMH 25698: C: Dorsal view, D: Ventral view. Scale bar C D 5 cm.

27 274 NF Simon Weigmann Figure 21 Neotrygon kuhlii: A B: ZMH 25699: A: Dorsal view, B: Ventral view. Scale bar A B 5 cm. C F: ZMH 25700: C D: Embryo, 94 mm disc width: C: Dorsal view, D: Ventral view. Scale bar C D 5 cm. E F: Embryo, 95 mm disc width: E: Dorsal view, F: Ventral view. Scale bar E F 5 cm.

28 Contribution to the taxonomy and distribution of eight ray species (Chondrichthyes, Batoidea) from coastal waters of Thailand , diplospondylous centra: ca. 96; pectoral fin rays (left/right): 112/112, ventral fin rays (left/right): 20/20 (excluding claspers). ZMH 25699: Vertebral centra anterior to the first sting: 101, total vertebral centra count: ca. 130, monospondylous centra: 39, diplospondylous centra: ca. 91; pectoral fin rays (left/right): 109/110, ventral fin rays (left/right): 24/23. ZMH 25700: Embryo, 94 mm wide: Vertebral centra anterior to the first sting: ca. 110, total vertebral centra count: ca Embryo, 95 mm wide: Vertebral centra anterior to the first sting: ca. 110, total vertebral centra count: ca Last & Stevens (2009) indicate a maximal disc width of at least 47 cm and a maximal total length of at least 76 cm as well as a birth size of cm disc width and a maturing size of about 29 cm (males) or 31 cm (females) width for Neotrygon kuhlii. With a disc width of 9.4 cm and 9.5 cm respectively the two examined embryos were obviously not close to birth. This is further evidenced by the not yet developed bluish spots in both embryos (Fig. 21, C and E). As typical for this species, the spots are present in their mother. The small male with 18.6 cm disc width (ZMH 25697) is still juvenile with not fully developed claspers (Fig. 20, B), whereas the large male with 32.4 cm disc width is mature with well-developed claspers (Fig. 20, D). Although Neotrygon kuhlii is very common and widespread (Last & Compagno 1999, Hennemann 2001, Daley et al. 2007, Last & Stevens 2009), there are still knowledge gaps regarding its biology and distribution. According to Debelius (1993) this species is always found far away from reefs, whereas it always stays nearby reefs after Patzner & Moosleitner (1999). Neotrygon kuhlii lives inshore to depths of 50 m (Debelius 1993, Patzner & Moosleitner 1999, Hennemann 2001) or 90 m respectively (Last & Compagno 1999, Last & Stevens 2009). By analyzing the catch data of 625 Australian expeditions (CSIRO 2004), of which 29 voyages produced catches of Neotrygon kuhlii with available depth data, the known depth distribution for Neotrygon kuhlii could be extended considerably. The detailed data about all catches of this species in depths deeper than 90 m are shown in Table 2. The considerable amount of catches below 90 m, as shown in Table 2, reveals that catches in such depths are not outliers. It can be seen that down to 110 m depth Neotrygon kuhlii was caught in quite high numbers, for example 10 specimens at station 33 (Cour 051) in 110 m depth and 11 exemplars at station 32 (also Cour 051) in 102 m depth. However, even down to 132 m depth there were quite many catches so that Neotrygon kuhlii seems to occur quite regularly in such depths. The specimen caught in 170 m depth represents a single value and is much deeper than all other catches. Therefore, such depths are probably not typical for the species. Despite its high abundance and wide distribution the known distribution area of Neotrygon kuhlii is partially fragmentary and varies between many references: According to Debelius (1993) the species is found in the whole Indian Ocean including the Red Sea. Following Burgess (1990) it occurs in the northern and western Indian Ocean, the Red Sea, the Great Barrier Reef and Southeast Asia. A similar distribution

29 276 NF Simon Weigmann Table 2 Voyages and stations with catches of Neotrygon kuhlii below 90 m depth. voyage station depth [m] number of specimens Cour not specified, at least 1 SO 5/ Cour SO 5/ Cour Cour not specified, at least 1 Cour SO 5/ Cour SO 5/ Cour not specified, at least 1 Cour not specified, at least 1 Cour SO 4/ not specified, at least 1 Cour Cour Cour SO 7/ total: At least 45 specimens area is mentioned by Patzner & Moosleitner (1999) and Hennemann (2001), who list a distribution from Durban (South Africa) over the Red Sea and Samoa until New Caledonia and Japan in the North. Last & Compagno (1999) expand this distribution area by Melanesia and Micronesia. Following Last & Stevens (2009) Neotrygon kuhlii is found in the coastal waters of South Africa as well as from West India over the whole northern Indian Ocean, Southeast Asia and northern Australia until Melanesia and Micronesia in the East and Japan and East China in the North. With the exception of three specimens of Neotrygon kuhlii that were reported from southern Africa (near Durban) by Compagno & Heemstra (1984) as well as five specimens from Zanzibar in the collection of the British Museum of Natural History, confirmed catches from the eastern coast of Africa and the Arabian coasts are still unknown. Therefore, more material from these regions would be desirable. The reports from the Red Sea should also be subject to review because contrary to many other authors Last & Stevens (2009) do not include the Red Sea in the distribution area of Neotrygon kuhlii. Improving the knowledge about Neotrygon kuhlii is complicated by the existence of different morphological variations, some of which probably not belonging to this species (Last & Compagno 1999, Last & Stevens 2009). Despite the uncertainties in biology and distribution, Neotrygon kuhlii belongs to

30 Contribution to the taxonomy and distribution of eight ray species (Chondrichthyes, Batoidea) from coastal waters of Thailand 277 Figure 22 Neotrygon kuhlii: Radiograph of specimen ZMH the most intensively commercially used ray species and is commercially fished in Thailand, too (Vidthayanon 2002, Last & Compagno 1999). However, exact data about the population status and possible negative trends in the populations of this species have not yet been collected (Fahmi & White 2007). Two habitus photographs of each of the five specimens (ZMH ZMH 25700) are shown in Figs. 20 and 21 and a radiograph of specimen ZMH in Fig. 22. The measurements of the five specimens can be found in Tables A15 A19. Aetomylaeus nichofii (Bloch & Schneider, 1801) The specimen of Aetomylaeus nichofii (ZMH 25701) was caught by local fishermen in the East Andaman Sea near Kantang on the 8 th December Like in all members of the family Myliobatidae as described by Last & Stevens (2009) the anterior part of the head of the examined specimen is rounded and extended beyond the disc and its eyes are located laterally on the sides of the head (Fig. 23, A B). Furthermore, it has a broad and lozenge-shaped disc and seven rows of plate-like teeth. As specified by Last & Stevens (2009) the examined specimen like all four species of the genus Aeto-

31 278 NF Simon Weigmann Figure 23 Aetomylaeus nichofii, ZMH 25701: A: Dorsal view, B: Ventral view. Scale bar A B 5 cm. mylaeus has no stinging spine, the single fleshy lobe around the snout is not connected to the pectoral fins, the internasal distance is shorter than the direct distance from the snout tip to the nostrils and the internasal flap is skirt-shaped (Fig. 23, A B). The mouth of the examined specimen is located two orbit diameters posterior to the snout tip, whereas following Last & Stevens (2009) it is three to four orbit diameters posterior to the snout tip in Aetomylaeus vespertilio (Bleeker, 1852). Contrary to Aetomylaeus maculatus (Gray, 1834) and A. vespertilio the origin of the dorsal fin is in front of the pelvic-fin insertions in the examined specimen (Fig. 23, A). In Aetomylaeus maculatus and A. vespertilio the dorsal fin originates posterior to the pelvic-fin insertions (Compagno & Last 1999a). The only other species of Aetomylaeus which has the dorsal fin origin anterior or opposite to the pelvic-fin insertions is A. milvus (Müller & Henle, 1841). However, this species has prominent ocelli on its dorsal surface, whereas the examined specimen has a plain dark brown color (Fig. 23, A) as described for adult Aetomylaeus nichofii specimens by Last & Stevens (2009). The juveniles of Aetomylaeus nichofii normally have about eight weak bands on the upper surface, but these bands probably faded during the storage in alcohol. Like Aetomylaeus milvus the species A. maculatus and A. vespertilio also have prominent color patterns on their upper surface. The specimen ZMH has seven tooth rows each in the upper and lower jaw as typical for the genus.

32 Contribution to the taxonomy and distribution of eight ray species (Chondrichthyes, Batoidea) from coastal waters of Thailand 279 Figure 24 Aetomylaeus nichofii: Radiograph of specimen ZMH Its internal meristics are as follows (Fig. 24): Vertebral centra anterior to the dorsal fin: 43, total vertebral centra count: 82, monospondylous centra: 36, diplospondylous centra: 46; pectoral fin rays (left/right): 89/88, ventral fin rays (left/right): 21/21. Last & Stevens (2009) indicate a maximal disc width of at least 72 cm and a maturity size of cm in males for Aetomylaeus nichofii. Therefore, the examined specimen is counted as juvenile with 31.5 cm disc width. Although Aetomylaeus nichofii is widespread, it is still a little known ray species. There is a small-scale commercial fishery for this species in parts of Asia. Its distribution ranges from at least India (probably the Arabian Gulf) over whole Southeast Asia until northern Australia in the South and South Japan in the North (Last & Stevens 2009). Compagno & Last (1999a) suppose that the real distribution area is even bigger and possibly includes South and East Africa as well as the Red Sea, but this assumption needs to be verified. Two habitus photographs of specimen ZMH are shown in Fig. 23 and a radiograph of the specimen in Fig. 24. Its measurements can be found in Table A20.

33 280 NF Simon Weigmann Rhinobatos formosensis Norman, 1926 The two specimens of Rhinobatos formosensis (ZMH and ZMH 25694) were caught between the 5 th and 11 th December 1993 in the coastal waters around Thailand, but unfortunately the exact location is unknown. As specified for the family Rhinobatidae by Compagno & Last (1999b) both specimens have a more or less shark-like body shape and the moderately large pectoral fins originate in front of the mouth and terminate posterior to the origins of the pelvic fins. Furthermore, following Last & Stevens (2009), all Rhinobatidae have a not well-developed ventral lobe of the caudal fin and the origin of the first dorsal fin is situated posterior to the pelvic fins (Fig. 27, A D). Like all members of the genus Rhinobatos the examined specimens have as described by Compagno & Last (1999b) no nasal curtain, diagonal nostrils (Fig. 27, B, D) and spiracles with a pair of narrow folds on their posterior margins (Fig. 27, A, C). The last substantial revision of the genus Rhinobatos was published by Norman (1926) when several species of the genus were not yet known. After ruling out all species described after Norman s revision, two species groups of quite similar species came into question for the determination of the two examined specimens: Species group 1 with R. holcorhynchus Norman, 1922, R. formosensis, and R. schlegelii Müller & Henle, 1841 as well as group 2 with R. hynnicephalus Richardson, 1846, R. rhinobatos (Linnaeus, 1758), R. annandalei Norman, 1926, and R. lionotus Norman, Important ratios for the differentiation between these groups are listed in Table 3. Table 3 Comparison of relevant ratios in the examined specimens and the two species groups. ratio ZMH ZMH species group 1 species group 2 POB/ISP almost 3 POR/MOW IDS/D1B less than 3 with POB = preorbital space, ISP = interspiracular distance, POR = preoral space, MOW = mouth width, IDS = interdorsal space, D1B = base length of first dorsal fin. Following these ratios the two examined specimens have to be assigned to species group 2. In this group the species Rhinobatos hynnicephalus and R. rhinobatos can be excluded because of the well-developed inner spiracular folds of the examined specimens (Fig. 27, A, C), whereas they are only small or rudimentary in R. hynnicephalus and R. rhinobatos. Contrary to Rhinobatos annandalei, which has a row of spines dorsomedially, both examined specimens only have a row of minute tubercles on the dorsal midline. Therefore, only Rhinobatos lionotus comes into question, which has a well-developed inner spiracular fold and a dorsal row of minute tubercles following Norman (1926). However, the snout of this species in the drawing by Norman (1926) is strikingly shorter than in the examined specimens (Fig. 25, A B).

34 Contribution to the taxonomy and distribution of eight ray species (Chondrichthyes, Batoidea) from coastal waters of Thailand 281 Figure 25 Dorsal heads: A: Rhinobatos lionotus (after Norman 1926), B: ZMH The reason for this discrepancy can be found out when carefully checking the other drawings of Norman (1926). The difference is caused by deviations in the definitions of some of the measurements. These deviations produce different values for some of the measurements between the definitions of Norman (1926) and Bigelow & Schroeder (1953). The differences are contrasted for both examined specimens in Table 4. Table 4 Important measurements for specimens ZMH and ZMH following Bigelow & Schroeder (1953) and Norman (1926). measurement ZMH ZMH value [mm] after BIGELOW &SCHROEDER (1953) value [mm] after NORMAN (1926) value [mm] after BIGELOW &SCHROEDER (1953) POR ISP D1B IDS MOW value [mm] after NORMAN (1926) When calculating the ratios with the values measured after Norman (1926) the values shown in Table 5 result. Table 5 Comparison of relevant ratios measured after Norman (1926) in the examined specimens and the two species groups. ratio ZMH ZMH species group 1 species group 2 POB/ISP almost 3 POR/MOW IDS/D1B less than 3

35 282 NF Simon Weigmann Figure 26 Dorsal heads: A: ZMH 25694, B D: Drawings after Norman (1926): B: Rhinobatos formosensis, C: R. schlegelii, D:R. holcorhynchus. The blue arrows point at the rostral ridges. With these readjusted values the two specimens ZMH and ZMH have to be assigned to species group 1 (with the values of the ratio preoral snout length to mouth width lying between both groups). In species group 1 both specimens can clearly be assigned to Rhinobatos formosensis because the formation of their rostral ridges is equivalent to that drawn by Norman (1926) (Fig. 26, A, B). In Rhinobatos schlegelli the rostral ridges run too close to each other (Fig. 26, C) and in R. holcorhynchus they are widely separated throughout their length and run together not until close to the snout tip (Fig. 26, D). Rhinobatos holcorhynchus can, furthermore, be excluded following Norman (1926) because of the ratio nostril length (NOW) to internarial distance (INW) being 1.8, whereas it is 1.33 to 1.5 in R. formosensis and R. schlegelii. In specimen ZMH the ratio NOW to INW is 1.47 and in ZMH it is Additionally, Rhinobatos holcorhynchus has a row of blunt tubercles in the midline of the back, whereas the tubercles are minute in both examined specimens (Fig. 27, A, C) as specified for R. formosensis and R. schlegelii by Norman (1926). Specimen ZMH has 103 tooth rows in the upper and 105 in the lower jaw, specimen ZMH has 79 rows in the upper and about 80 rows in the lower jaw. The internal meristics are as follows: ZMH (Fig. 28): Synarcual with an anterior centrum-free region of 14 segments (SYS) and a posterior region with 11 embedded centra (SYC); total vertebral centra count: 190, total free vertebral centra count: 179, monospondylous centra (including SYC): 37, diplospondylous centra: 153, total vertebral centra anterior to the first dorsal fin: 68, total vertebral centra anterior to the second dorsal fin: 111, total vertebral centra anterior to the caudal fin: 146, vertebral centra in the caudal fin: 44; pectoral fin rays (left/right): propterygials: 33/33, mesopterygials: 6/7, neopterygials: 1/2, metapterygials: 27/28, total basal radials: 67/ 70; ventral fin rays (left/right): 27/25 (excluding claspers). ZMH 25694: SYS: 13 seg-

36 Contribution to the taxonomy and distribution of eight ray species (Chondrichthyes, Batoidea) from coastal waters of Thailand 283 Figure 27 Rhinobatos formosensis: A B: ZMH 25693: A: Dorsal view, B: Ventral view. Scale bar A B 5 cm. C D: ZMH 25694: C: Dorsal view, D: Ventral view. Scale bar C D 5 cm.

37 284 NF Simon Weigmann Figure 28 Rhinobatos formosensis: Radiograph of specimen ZMH ments, SYC: 13 centra; total vertebral centra count: 187, total free vertebral centra count: 174, monospondylous centra (including SYC): 37, diplospondylous centra: 150, total vertebral centra anterior to the first dorsal fin: 70, total vertebral centra anterior to the second dorsal fin: 110, total vertebral centra anterior to the caudal fin: 146, vertebral centra in the caudal fin: 41; pectoral fin rays (left/right): propterygials: 33/33, mesopterygials: 7/7, neopterygials: 2/2, metapterygials: 26/26, total basal radials: 68/ 68; ventral fin rays (left/right): 25/25 (excluding claspers). Following Compagno & Last (1999b) Rhinobatos formosensis reaches a maximal total length of at least 63 cm. The adult male specimen (ZMH 25693) with fully developed claspers (Fig. 27, B) is considerably larger than this size with a total length of 73 cm. As this species gets mature not until 56 cm total length (Compagno & Last 1999b), more than 63 cm total length can surely be expected compared to the maturity and maximal length of other species of the genus. The specimen ZMH with a total length of 42 cm is clearly juvenile due to the hardly developed claspers (Fig. 27, D). The biology and fishery exploitation of Rhinobatos formosensis are almost unknown (Compagno & Last 1999b). Compagno & Last (1999b) list a depth distribution from the intertidal zone to 119 m depth and a very limited distribution area because the species

38 Contribution to the taxonomy and distribution of eight ray species (Chondrichthyes, Batoidea) from coastal waters of Thailand 285 Figure 29 Rhinobatos formosensis: Distribution area after Last & Compagno (1999) and GBIF (2010), with the possible catch location of specimens ZMH and ZMH in blue and white stripes. has been known only from Taiwan and the mouth of Manila Bay, Luzon, Philippines. Therefore, it was named Taiwan Guitarfish. The two catches from Thailand represent a considerable enlargement of the distribution area and prove that this species is not endemic to the Taiwanese region. The new records are about 2500 kilometers away from the formerly known distribution area. Two habitus photographs of each of the specimens ZMH and ZMH are shown in Fig. 27 and a radiograph of specimen ZMH in Fig. 28. A distribution map for Rhinobatos formosensis is pictured in Fig. 29, in which the distribution area after Last & Compagno (1999) as well as the species dataset of the Global Biodiversity Information Facility (GBIF 2010) is marked in red and white stripes and the possible catch location of the two specimens ZMH and ZMH as blue and white stripes. The measurements of both specimens can be found in Tables A21 A22. Rhynchobatus australiae Whitley, 1939 The two specimens of Rhynchobatus australiae (ZMH and ZMH 25688) were caught by local fishermen in the Gulf of Thailand near Pak Phanang on the 7 th December As members of the monotypic family Rhynchobatidae (wedgefishes) and consequently the genus Rhynchobatus they are following Last & Stevens (2009)

39 286 NF Simon Weigmann shark-like rays with the pectoral fins and head distinct from each other as well as a caudal fin with both lobes well-developed and a deeply concave posterior margin (Figs ). Only for Rhynchobatus djiddensis (Forsskål, 1775) a short ventral caudal fin lobe was described and pictured by Raje et al. (2007). This abnormality was probably caused by a healed lesion because no other such specimens have been described in literature and the well-developed lower caudal fin lobe is even an important character of the whole family Rhynchobatidae. Further characteristics of this family as described by Compagno & Last (1999c) are a line of small thorns dorsomedially and two short lines on the shoulders, the orbits and sometimes the snout as well as angular and low pectoral and pelvic fins (Figs ). Until now six species of the genus Rhynchobatus have been described. Rhynchobatus laevis (Bloch & Schneider, 1801) can be excluded in the determination of the examined specimens because in this species the origin of the first dorsal fin is situated opposite to the pelvic fins (Last & Stevens 2009), whereas in the examined specimens it is located posterior to the pelvic fins (Fig. 31, C, D). Furthermore, Rhynchobatus laevis has a small or no black pectoral marking with only three sorrounding white spots (Last & Stevens 2009), whereas it is prominent and surrounded by four white spots in the examined specimens (Figs ). Rhynchobatus luebberti Ehrenbaum, 1915 can also be excluded because it has no black pectoral markings, but, instead, black blotches on both sides of the median dorsal row of tubercles. Additionally, this species only occurs in the tropical East Atlantic Ocean (Stehmann 1981). The two examined specimens, furthermore, do not match the descriptions of Rhynchobatus palpebratus Compagno & Last, 2008 und R. springeri Last, White & Pogonoski, 2010 because they have no dark eyebrow-like markings above the eyes and also no dark spot posterior to the orbit. Rhynchobatus palpebratus and R. springeri always have eyebrow-like markings and often a dark spot posterior to the orbit (Last & Stevens 2009, Last et al. 2010). Fig. 30, which shows the two examined specimens directly after the catch, proves that the examined specimens really did not have any eyebrow-like markings, which otherwise could eventually have gotten lost during the storage in alcohol. Furthermore, contrary to Rhynchobatus palpebratus and R. springeri, the origin of the first dorsal fin is situated more posteriorly in the two examined specimens: In Rhynchobatus palpebratus the origin of the first dorsal fin is over or slightly anterior to the pelvic fin origins (Compagno & Last 2008), in R. springeri the origin of the first dorsal fin is over the origin of the pelvic fin bases (Last et al. 2010). In the examined specimens the first dorsal fin is situated posterior to the pelvic fin origins (Fig. 31, C, D) as described for Rhynchobatus australiae by Last & Stevens (2009) and for R. djiddensis by Wallace (1967). The examined specimens (ZMH and ZMH 25688) morphologically correspond with the descriptions of the species Rhynchobatus djiddensis (e. g. Day 1878,

40 Contribution to the taxonomy and distribution of eight ray species (Chondrichthyes, Batoidea) from coastal waters of Thailand 287 Figure 30 Photograph of the two specimens ZMH and directly after the catch on the fish market in Pak Phanang by Matthias Stehmann. Day 1889, Jordan & Fowler 1903, Garman 1913, Monkolprasit 1984 and Smith & Heemstra 2003) and Rhynchobatus australiae (e. g. Compagno & Last 1999c, White et al. 2006, Last & Stevens 2009, Last et al. 2010). Both species are morphologically extremely similar and are mainly differentiated by means of their distribution area: Rhynchobatus djiddensis occurs in the West and R. australiae in the East Indian Ocean (Compagno & Last 1999c). The main determination problem is that until few years ago all wedgefishes worldwide were applied to Rhynchobatus djiddensis (Last & Stevens 2009). That means that probably at least some of the available, above listed descriptions of Rhynchobatus djiddensis describe specimens of R. australiae or maybe even other similar, recently described species of the genus. Due to this problem the exact distribution areas of the different Rhynchobatus species are still unknown, too. Day (1878, 1889), Jordan & Fowler (1903), Garman (1913), Monkolprasit (1984) and Smith & Heemstra (2003) unfortunately do not mention the origin of the Rhynchobatus djiddensis specimens described by them. With regard to the generally fewer expeditions to the West than to the East Indian Ocean it is likely that the majority of the descriptions of Rhynchobatus djiddensis are based on specimens from the East Indian Ocean and, therefore, on R. australiae or other similar species and not the real R. djiddensis. The only reliable source for a description of surely real Rhynchobatus djiddensis specimens seems to be Wallace (1967) who studied 68 specimens of both sexes, from embryos to adults, from many locations at the Southeast African Natal coast between Port Edward and Sinkwazi. Regarding the amount of dorsal tubercles there are differ-

41 288 NF Simon Weigmann ences between the description by Wallace (1967) and the two specimens ZMH and ZMH 25688: According to Wallace (1967) an inner series of 7 16 tubercles and an outer series of 2 3 tubercles are present above each shoulder in Rhynchobatus djiddensis, whereas each of the examined specimens has 6 inner and 5 outer tubercles. Additionally, there are differences in the amounts of median tubercles: Following Wallace (1967) Rhynchobatus djiddensis has a row of 29 to 50 tubercles anterior to the first dorsal fin and a similar row of 31 to 48 tubercles on the interdorsal space. The examined specimens each have 29 to 30 tubercles anterior to the first dorsal fin and a row of only 17 to 18 tubercles on the interdorsal space. The differences are probably not caused by the juvenility of the two specimens ZMH and ZMH because Wallace (1967) examined individuals of all stages of growth. However, it remains unclear if the counts by Wallace (1967) are based on all 68 mentioned specimens or on few specimens only as well as how variable this character generally is in Rhynchobatus djiddensis. Therefore, despite the possible differences in the amounts of dorsal tubercles, certain morphological differences between Rhynchobatus djiddensis and R. australiae remain unavailable because both species generally have not been differentiated clearly until recently. Another possible difference between Rhynchobatus australiae and R. djiddensis arises from the tooth row counts: Many references specify tooth row counts of about 40 rows per jaw for Rhynchobatus djiddensis (40 42 rows per jaw according to Day (1878), Day (1889), Jordan & Fowler (1903); about 40 rows per jaw following Monkolprasit (1984); rows in the upper and rows in the lower jaw after Wallace (1967), Smith & Heemstra (2003)). These counts are far lower than the counts of the two specimens ZMH and ZMH as well as those of the other similar species of the genus, which all have more than 50 tooth rows per jaw. Specimen ZMH has 60 tooth rows in the upper and 58 in the lower jaw, ZMH has 64 rows in the upper and 62 rows in the lower jaw, Rhynchobatus palpebratus has about 52 tooth rows in the upper jaw (Compagno & Last (2008)) and R. springeri has about 52 rows in the upper jaw, too (Last et al. (2010)). The upper tooth row count of 27 listed by Fowler et al. (1941) for Rhynchobatus yentinensis Wang, (1933) is obviously not correct because it is extremely low compared to the other species of the genus. Rhynchobatus yentinensis is considered to be synonymous with R. laevis by Compagno & Last (2008) and Eschmeyer (2010). As all morphologically quite similar Rhynchobatus species with the exception of R. djiddensis have similar amounts of tooth rows in their jaws, it is unlikely that R. djiddensis really has much less tooth rows per jaw. The paper by Wallace (1967) indicates that the low counts for Rhynchobatus djiddensis are really not correct because the lower jaw of a Rhynchobatus djiddensis pictured on a photograph in Wallace (1967) has about 70 tooth rows, although Wallace lists round about 40 rows per jaw. The counts by Smith & Heemstra (2003) are exactly identical with those by Wallace (1967) and

42 Contribution to the taxonomy and distribution of eight ray species (Chondrichthyes, Batoidea) from coastal waters of Thailand 289 Figure 31 Rhynchobatus australiae: A C: ZMH 25687: A: Lateral view, B: Ventral view, C: Dorsal view. Scale bar A C 5 cm. D: ZMH 25688: Dorsal view. Scale bar D 5 cm.

43 290 NF Simon Weigmann Figure 32 Rhynchobatus australiae: A B: ZMH 25688: A: Ventral view, B: Lateral view. Scale bar A B 5 cm. thus are obviously based on Wallace s numbers. The counts by Day (1878, 1889) and Jordan & Fowler (1903) are also identical among each other and, therefore, are probably based on the same source. The possible reason for the low tooth row counts is a wrong method of counting the rows by not counting in a zigzag pattern but, instead, only counting the parallel transverse rows, which results in much lower counts than in reality. The correct method of counting the tooth rows of rays is shown in Fig. 2. The same guess is pointed out by Krefft (1961): In the remarks on Himantura gerrardi Krefft expresses the assumption that Fowler (1941) possibly made a mistake when counting the tooth rows of this species due to counting transverse rows instead of a zigzag pattern because Krefft got similar counts as Fowler when recounting the teeth of this species with the wrong counting method. For Rhynchobatus djiddensis Krefft (1961) lists realistic tooth row counts with 57 tooth rows in the upper and 59 rows in the lower jaw. Additionally, Krefft (1961) mentions the low counts by Fowler (1941) also for this species, but does not discuss possible reasons for the difference. Krefft (1961) probably expects the same reason as which he had expressed for the low counts in Himantura gerrardi. Garman (1913) also gives realistic counts for Rhynchobatus djiddensis with 63 tooth rows for the upper jaw. Again, at least some of the counts could represent counts of Rhynchobatus australiae or other similar species in reality due to the not known origin of the examined specimens.

44 Contribution to the taxonomy and distribution of eight ray species (Chondrichthyes, Batoidea) from coastal waters of Thailand 291 Figure 33 Rhynchobatus australiae: Radiograph of specimen ZMH The internal meristics are as follows: ZMH (Fig. 33): SYS: 19 segments, SYC: 12 centra; total vertebral centra count: 168, total free vertebral centra count: 156, monospondylous centra (including SYC): 44, diplospondylous centra: 124, total vertebral centra anterior to the first dorsal fin: 39, total vertebral centra anterior to the second dorsal fin: 89, total vertebral centra anterior to the caudal fin: 130, vertebral centra in the caudal fin: 38; pectoral fin rays (left/right): free radials before propterygium: 5/5, propterygials: 25/25, mesopterygials: 4/4, neopterygials: 4/5, metapterygials: 28/26, total basal radials: 66/65; ventral fin rays (left/right): ca. 25/ca. 25 (excluding claspers). ZMH 25688: SYS: 20 segments, SYC: 11 centra; total vertebral centra count: 175, total free vertebral centra count: 164, monospondylous centra (including SYC): 44, diplospondylous centra: 131, total vertebral centra anterior to the first dorsal fin: 40, total vertebral centra anterior to the second dorsal fin: 94, total vertebral centra anterior to the caudal fin: 133, vertebral centra in the caudal fin: 42; pectoral fin rays (left/right): free radials before propterygium: 6/7, propterygials: 26/25, mesopterygials: 4/4, neopterygials: 4/4, metapterygials: 26/25, total basal radials: 66/65; ventral fin rays (left/right): 27/27. As in cases of the tooth rows, the vertebral counts for Rhynchobatus djiddensis are